Project Summary/Abstract Studying how cortical networks control coordinated limb movements advances our understanding of the principles of biological information processing; this knowledge has direct applications in the field of brain-computer interfaces (BCIs) for people with motor disorders. Neural activity in motor cortex is usually interpreted as a command signal that drives movement using a static one-to-one mapping. However, recent findings suggest that the relationship between neural activity and movement is not fixed, but instead changes depending on context. Here, we develop a new model based on the idea that motor cortex dynamically encodes the interactions between the body and the environment. Under this formulation, a specific movement is associated with many different activity patterns when performed in different behavioral settings, matching recent experimental findings in humans and other primates. Based on this theory, we propose gathering information about the environment through artificial vision, and using this information as a filter to interpret motor cortical activity more accurately. This holistic approach will allow us understand the role of different cortical areas driving behavior in more detail, and optimize our strategies for extracting movement-related information from cortical circuits. This project will provide a theoretical framework to explore the possibilities of combining biological signals with easily accessible artificial data streams to solve complex problems such as motor control. Our results will drive new advances in neuroengineering, including the development of hybrid control systems for dexterous robotic limbs. !